의학물리학

본문글자크기
  • [Radiother Oncol.] Clinical evaluation of atlas- and deep learning-based automatic segmentation of multiple organs and clinical target volumes for breast cancer

    [Radiother Oncol.] Clinical evaluation of atlas- and deep learning-based automatic segmentation of multiple organs and clinical target volumes for breast cancer

    연세의대 / 최민서, 장지석*, 김진성*

  • 출처
    Radiother Oncol.
  • 등재일
    2020 Sep 26
  • 저널이슈번호
    S0167-8140(20)30820-3. doi: 10.1016/j.radonc.2020.09.045.
  • 내용

    바로가기  >

    Abstract
    Manual segmentation is the gold standard method for radiation therapy planning; however, it is time-consuming and prone to inter- and intra-observer variation, giving rise to interests in auto-segmentation methods. We evaluated the feasibility of deep learning-based auto-segmentation (DLBAS) in comparison to commercially available atlas-based segmentation solutions (ABAS) for breast cancer radiation therapy. This study used contrast-enhanced planning computed tomography scans from 62 patients with breast cancer who underwent breast-conservation surgery. Contours of target volumes (CTVs), organs, and heart substructures were generated using two commercial ABAS solutions and DLBAS using fully convolutional DenseNet. The accuracy of the segmentation was assessed using 14 test patients using the Dice Similarity Coefficient and Hausdorff Distance referencing the expert contours. A sensitivity analysis was performed using non-contrast planning CT from 14 additional patients. Compared to ABAS, the proposed DLBAS model yielded more consistent results and the highest average Dice Similarity Coefficient values and lowest Hausdorff Distances, especially for CTVs and the substructures of the heart. ABAS showed limited performance in soft-tissue-based regions, such as the esophagus, cardiac arteries, and smaller CTVs. The results of sensitivity analysis between contrast and non-contrast CT test sets showed little difference in the performance of DLBAS and conversely, a large discrepancy for ABAS. The proposed DLBAS algorithm was more consistent and robust in its performance than ABAS across the majority of structures when examining both CTVs and normal organs. DLBAS has great potential to aid a key process in the radiation therapy workflow, helping optimise and reduce the clinical workload.

     

    Affiliations

    Min Seo Choi  1 , Byeong Su Choi  1 , Seung Yeun Chung  1 , Nalee Kim  1 , Jaehee Chun  1 , Yong Bae Kim  1 , Jee Suk Chang  2 , Jin Sung Kim  3
    1 Department of Radiation Oncology, Yonsei Cancer Center, Yonsei University College of Medicine, 50-1 Yonsei-ro, Seodaemoon-gu, Seoul 03722, South Korea.
    2 Department of Radiation Oncology, Yonsei Cancer Center, Yonsei University College of Medicine, 50-1 Yonsei-ro, Seodaemoon-gu, Seoul 03722, South Korea. Electronic address: changjeesuk@yuhs.ac.
    3 Department of Radiation Oncology, Yonsei Cancer Center, Yonsei University College of Medicine, 50-1 Yonsei-ro, Seodaemoon-gu, Seoul 03722, South Korea. Electronic address: jinsung@yuhs.ac.

  • 키워드
    Artificial Intelligence; Breast cancer; Clinical Target Volume; Commercial Atlas-based autosegmentation; Deep learning-based autosegmentation; Organs at risk; Radiation Therapy.
  • 연구소개
    최근 많은 양의 반복적인 작업과 관찰자 간의 편차를 줄이기 위하여 방사선 치료 계획의 핵심 단계인 contouring을 자동화하려는 노력들이 많이 이루어지고 있습니다. 저희 연구 논문에서는 최근 각광을 받고 있는 딥러닝 기반의 자동 컨투어링과 기존의 아틀라스 방식의 자동 contouring 알고리즘을 유방암 환자 데이터를 이용하여 비교했습니다. 각 알고리즘의 성능을 심장 세부 구조를 포함한 정상 장기들과 손상 위험 장기들에서 상대적으로 비교하여 딥러닝 알고리즘이 아틀라스 알고리즘에 비해 더 정확하고 robust한 결과를 제공한다는 것을 검증하였습니다. 또한 본 연구 논문에서 딥러닝 기반의 자동 contouring 알고리즘의 임상적 실용성과 실행될 만한 가치가 있는지에 대해 논의했기 때문에 향후 각 기관에서 decision making을 하는데 도움이 될만한 정보라고 생각합니다.
  • 편집위원

    AI를 이용한 오토 segmentation이 앞으로 널리 사용될 수 밖에 없을텐데, 현재는 간단한 정상장기를 그리는 정도이지만, 앞으로는 본 논문에서 연구한 바 와 같이 치료체적까지 많은 도움을 받을 것으로 예상된다.

    2020-11-10 09:54:46

  • 덧글달기
    덧글달기
       IP : 3.89.200.155

    등록